Mailbook: A social network against spamming

Spam is the main problem of email systems nowadays. The total amount of spam emails account for more than 75% of the total emails exchanged worldwide; recent reports raise this number up to more than 90%. Novel anti-spam solutions are proposed constantly, to be followed by announcements of sophisticated methods to overcome them through the use of advanced software to reach the spammers´ goal. In this paper we propose a collaborative spam filter over a social network, exchanging vote databases containing the hash values of the emails perceived as spam by its users, the mailbook. Social networks are blooming nowadays and users are accustomed to their use more and more every day. Our proposal builds upon that strong attachment between friends and people with the same interests and habits. We propose a user based collaborative approach to address the spam problem. Users characterize spam mail and exchange their votes among their friends through mailbook. User profiles are created in mailbook to express user interests, which in turn are used for evaluating mail as spam according to the user´s characteristics. Users also form groups of interests which are also used by our method as another mean to evaluate spam for the specific group in a more effective way.